[v2] drm: Check actual format for legacy pageflip.

Commit Message

Bas Nieuwenhuizen Jan. 10, 2021, 2:11 a.m. UTC

With modifiers one can actually have different format_info structs
for the same format, which now matters for AMDGPU since we convert
implicit modifiers to explicit modifiers with multiple planes.

I checked other drivers and it doesn't look like they end up triggering
this case so I think this is safe to relax.

Signed-off-by: Bas Nieuwenhuizen <[e-mail]>
Reviewed-by: Daniel Vetter <[e-mail]>
Reviewed-by: Zhan Liu <[e-mail]>
Acked-by: Christian König <[e-mail]>
Acked-by: Alex Deucher <[e-mail]>
Fixes: 816853f9dc40 ("drm/amd/display: Set new format info for converted metadata.")
---
 drivers/gpu/drm/drm_plane.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

Comments

Alex Deucher Jan. 11, 2021, 3:02 p.m. UTC | #1

On Sat, Jan 9, 2021 at 9:11 PM Bas Nieuwenhuizen
<[e-mail]> wrote:
>
> With modifiers one can actually have different format_info structs
> for the same format, which now matters for AMDGPU since we convert
> implicit modifiers to explicit modifiers with multiple planes.
>
> I checked other drivers and it doesn't look like they end up triggering
> this case so I think this is safe to relax.
>
> Signed-off-by: Bas Nieuwenhuizen <[e-mail]>
> Reviewed-by: Daniel Vetter <[e-mail]>
> Reviewed-by: Zhan Liu <[e-mail]>
> Acked-by: Christian König <[e-mail]>
> Acked-by: Alex Deucher <[e-mail]>
> Fixes: 816853f9dc40 ("drm/amd/display: Set new format info for converted metadata.")

Do you have commit rights to drm-misc or do you need someone to commit
this for you?

Thanks!

Alex

> ---
>  drivers/gpu/drm/drm_plane.c | 9 ++++++++-
>  1 file changed, 8 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/drm_plane.c b/drivers/gpu/drm/drm_plane.c
> index e6231947f987..a0cb746bcb0a 100644
> --- a/drivers/gpu/drm/drm_plane.c
> +++ b/drivers/gpu/drm/drm_plane.c
> @@ -1163,7 +1163,14 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
>         if (ret)
>                 goto out;
>
> -       if (old_fb->format != fb->format) {
> +       /*
> +        * Only check the FOURCC format code, excluding modifiers. This is
> +        * enough for all legacy drivers. Atomic drivers have their own
> +        * checks in their ->atomic_check implementation, which will
> +        * return -EINVAL if any hw or driver constraint is violated due
> +        * to modifier changes.
> +        */
> +       if (old_fb->format->format != fb->format->format) {
>                 DRM_DEBUG_KMS("Page flip is not allowed to change frame buffer format.\n");
>                 ret = -EINVAL;
>                 goto out;
> --
> 2.29.2
>
> _______________________________________________
> amd-gfx mailing list
> [e-mail]
> https://lists.freedesktop.org/mailman/listinfo/amd-gfx

Bas Nieuwenhuizen Jan. 11, 2021, 4:40 p.m. UTC | #2

On Mon, Jan 11, 2021 at 4:02 PM Alex Deucher <[e-mail]> wrote:
>
> On Sat, Jan 9, 2021 at 9:11 PM Bas Nieuwenhuizen
> <[e-mail]> wrote:
> >
> > With modifiers one can actually have different format_info structs
> > for the same format, which now matters for AMDGPU since we convert
> > implicit modifiers to explicit modifiers with multiple planes.
> >
> > I checked other drivers and it doesn't look like they end up triggering
> > this case so I think this is safe to relax.
> >
> > Signed-off-by: Bas Nieuwenhuizen <[e-mail]>
> > Reviewed-by: Daniel Vetter <[e-mail]>
> > Reviewed-by: Zhan Liu <[e-mail]>
> > Acked-by: Christian König <[e-mail]>
> > Acked-by: Alex Deucher <[e-mail]>
> > Fixes: 816853f9dc40 ("drm/amd/display: Set new format info for converted metadata.")
>
> Do you have commit rights to drm-misc or do you need someone to commit
> this for you?

I don't have commit rights so if the patch could be committed for me
that would be appreciated!
>
> Thanks!
>
> Alex
>
> > ---
> >  drivers/gpu/drm/drm_plane.c | 9 ++++++++-
> >  1 file changed, 8 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/gpu/drm/drm_plane.c b/drivers/gpu/drm/drm_plane.c
> > index e6231947f987..a0cb746bcb0a 100644
> > --- a/drivers/gpu/drm/drm_plane.c
> > +++ b/drivers/gpu/drm/drm_plane.c
> > @@ -1163,7 +1163,14 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
> >         if (ret)
> >                 goto out;
> >
> > -       if (old_fb->format != fb->format) {
> > +       /*
> > +        * Only check the FOURCC format code, excluding modifiers. This is
> > +        * enough for all legacy drivers. Atomic drivers have their own
> > +        * checks in their ->atomic_check implementation, which will
> > +        * return -EINVAL if any hw or driver constraint is violated due
> > +        * to modifier changes.
> > +        */
> > +       if (old_fb->format->format != fb->format->format) {
> >                 DRM_DEBUG_KMS("Page flip is not allowed to change frame buffer format.\n");
> >                 ret = -EINVAL;
> >                 goto out;
> > --
> > 2.29.2
> >
> > _______________________________________________
> > amd-gfx mailing list
> > [e-mail]
> > https://lists.freedesktop.org/mailman/listinfo/amd-gfx

Alex Deucher Jan. 11, 2021, 9:28 p.m. UTC | #3

On Mon, Jan 11, 2021 at 11:39 AM Bas Nieuwenhuizen
<[e-mail]> wrote:
>
> On Mon, Jan 11, 2021 at 4:02 PM Alex Deucher <[e-mail]> wrote:
> >
> > On Sat, Jan 9, 2021 at 9:11 PM Bas Nieuwenhuizen
> > <[e-mail]> wrote:
> > >
> > > With modifiers one can actually have different format_info structs
> > > for the same format, which now matters for AMDGPU since we convert
> > > implicit modifiers to explicit modifiers with multiple planes.
> > >
> > > I checked other drivers and it doesn't look like they end up triggering
> > > this case so I think this is safe to relax.
> > >
> > > Signed-off-by: Bas Nieuwenhuizen <[e-mail]>
> > > Reviewed-by: Daniel Vetter <[e-mail]>
> > > Reviewed-by: Zhan Liu <[e-mail]>
> > > Acked-by: Christian König <[e-mail]>
> > > Acked-by: Alex Deucher <[e-mail]>
> > > Fixes: 816853f9dc40 ("drm/amd/display: Set new format info for converted metadata.")
> >
> > Do you have commit rights to drm-misc or do you need someone to commit
> > this for you?
>
> I don't have commit rights so if the patch could be committed for me
> that would be appreciated!

Pushed to drm-misc-fixes.  Thanks!

If you want access to drm-misc, I don't see any reason you shouldn't have it.

Alex


> >
> > Thanks!
> >
> > Alex
> >
> > > ---
> > >  drivers/gpu/drm/drm_plane.c | 9 ++++++++-
> > >  1 file changed, 8 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/gpu/drm/drm_plane.c b/drivers/gpu/drm/drm_plane.c
> > > index e6231947f987..a0cb746bcb0a 100644
> > > --- a/drivers/gpu/drm/drm_plane.c
> > > +++ b/drivers/gpu/drm/drm_plane.c
> > > @@ -1163,7 +1163,14 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
> > >         if (ret)
> > >                 goto out;
> > >
> > > -       if (old_fb->format != fb->format) {
> > > +       /*
> > > +        * Only check the FOURCC format code, excluding modifiers. This is
> > > +        * enough for all legacy drivers. Atomic drivers have their own
> > > +        * checks in their ->atomic_check implementation, which will
> > > +        * return -EINVAL if any hw or driver constraint is violated due
> > > +        * to modifier changes.
> > > +        */
> > > +       if (old_fb->format->format != fb->format->format) {
> > >                 DRM_DEBUG_KMS("Page flip is not allowed to change frame buffer format.\n");
> > >                 ret = -EINVAL;
> > >                 goto out;
> > > --
> > > 2.29.2
> > >
> > > _______________________________________________
> > > amd-gfx mailing list
> > > [e-mail]
> > > https://lists.freedesktop.org/mailman/listinfo/amd-gfx

Daniel Vetter Jan. 12, 2021, 9:18 a.m. UTC | #4

On Mon, Jan 11, 2021 at 04:28:31PM -0500, Alex Deucher wrote:
> On Mon, Jan 11, 2021 at 11:39 AM Bas Nieuwenhuizen
> <[e-mail]> wrote:
> >
> > On Mon, Jan 11, 2021 at 4:02 PM Alex Deucher <[e-mail]> wrote:
> > >
> > > On Sat, Jan 9, 2021 at 9:11 PM Bas Nieuwenhuizen
> > > <[e-mail]> wrote:
> > > >
> > > > With modifiers one can actually have different format_info structs
> > > > for the same format, which now matters for AMDGPU since we convert
> > > > implicit modifiers to explicit modifiers with multiple planes.
> > > >
> > > > I checked other drivers and it doesn't look like they end up triggering
> > > > this case so I think this is safe to relax.
> > > >
> > > > Signed-off-by: Bas Nieuwenhuizen <[e-mail]>
> > > > Reviewed-by: Daniel Vetter <[e-mail]>
> > > > Reviewed-by: Zhan Liu <[e-mail]>
> > > > Acked-by: Christian König <[e-mail]>
> > > > Acked-by: Alex Deucher <[e-mail]>
> > > > Fixes: 816853f9dc40 ("drm/amd/display: Set new format info for converted metadata.")
> > >
> > > Do you have commit rights to drm-misc or do you need someone to commit
> > > this for you?
> >
> > I don't have commit rights so if the patch could be committed for me
> > that would be appreciated!
> 
> Pushed to drm-misc-fixes.  Thanks!
> 
> If you want access to drm-misc, I don't see any reason you shouldn't have it.

There's some old-school bash tooling involved since we're (not yet, I can
hope) doing gitlab MR:

https://drm.pages.freedesktop.org/maintainer-tools/getting-started.html

Otherwise makes sense imo.
-Daniel

> 
> Alex
> 
> 
> > >
> > > Thanks!
> > >
> > > Alex
> > >
> > > > ---
> > > >  drivers/gpu/drm/drm_plane.c | 9 ++++++++-
> > > >  1 file changed, 8 insertions(+), 1 deletion(-)
> > > >
> > > > diff --git a/drivers/gpu/drm/drm_plane.c b/drivers/gpu/drm/drm_plane.c
> > > > index e6231947f987..a0cb746bcb0a 100644
> > > > --- a/drivers/gpu/drm/drm_plane.c
> > > > +++ b/drivers/gpu/drm/drm_plane.c
> > > > @@ -1163,7 +1163,14 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
> > > >         if (ret)
> > > >                 goto out;
> > > >
> > > > -       if (old_fb->format != fb->format) {
> > > > +       /*
> > > > +        * Only check the FOURCC format code, excluding modifiers. This is
> > > > +        * enough for all legacy drivers. Atomic drivers have their own
> > > > +        * checks in their ->atomic_check implementation, which will
> > > > +        * return -EINVAL if any hw or driver constraint is violated due
> > > > +        * to modifier changes.
> > > > +        */
> > > > +       if (old_fb->format->format != fb->format->format) {
> > > >                 DRM_DEBUG_KMS("Page flip is not allowed to change frame buffer format.\n");
> > > >                 ret = -EINVAL;
> > > >                 goto out;
> > > > --
> > > > 2.29.2
> > > >
> > > > _______________________________________________
> > > > amd-gfx mailing list
> > > > [e-mail]
> > > > https://lists.freedesktop.org/mailman/listinfo/amd-gfx

Patch

diff --git a/drivers/gpu/drm/drm_plane.c b/drivers/gpu/drm/drm_plane.c
index e6231947f987..a0cb746bcb0a 100644
--- a/drivers/gpu/drm/drm_plane.c
+++ b/drivers/gpu/drm/drm_plane.c
@@ -1163,7 +1163,14 @@  int drm_mode_page_flip_ioctl(struct drm_device *dev,
 	if (ret)
 		goto out;
 
-	if (old_fb->format != fb->format) {
+	/*
+	 * Only check the FOURCC format code, excluding modifiers. This is
+	 * enough for all legacy drivers. Atomic drivers have their own
+	 * checks in their ->atomic_check implementation, which will
+	 * return -EINVAL if any hw or driver constraint is violated due
+	 * to modifier changes.
+	 */
+	if (old_fb->format->format != fb->format->format) {
 		DRM_DEBUG_KMS("Page flip is not allowed to change frame buffer format.\n");
 		ret = -EINVAL;
 		goto out;